Some platform videogames have this special level where the movement controls were reversed as a challenge to the player. The left becomes right, the up is now down. Anyone playing this level for the first time might fail the level several times, but it becomes easier when you later encounter it in other videogames. Super Karoshi and This is the only level are two examples that used that trope.
Matthew Cain (Duke University) and company published their findings in Attention, Perception and Psychophysics.
Video game expertise has been shown to have beneficial effects for visual attention processes, but the effects of action video game playing on executive functions, such as task switching and filtering out distracting information, are less well understood. In the main experiment presented here, video game players (VGPs) and nonplayers (nVGPs) switched between two tasks of unequal familiarity: a familiar task of responding in the direction indicated by an arrow, and a novel task of responding in the opposite direction. nVGPs had large response time costs for switching from the novel task to the familiar task, and small costs for switching from the familiar task to the novel task, replicating prior findings. However, as compared to the nVGPs, VGPs were more facile in switching between tasks, producing overall smaller and more symmetric switching costs, suggesting that experience with action video games produces improvements in executive functioning. In contrast, VGPs and nVGPs did not differ in filtering out the irrelevant flanking stimuli or in remembering details of aurally presented stories. The lack of global differences between the groups suggests that the improved task-switching performance seen in VGPs was not due to differences in global factors, such as VGPs being more motivated than nVGPs.
If I were a pokemon, this line of research would be my weakness. It took me a long time to review this article.
The authors examined videogame experience on three cognitive abilities: task switching, distractor filtering and logical memory. Task switching, using the videogame example mentioned earlier, involves switching between two dissimilar tasks. The cost of switching is response time where switching makes you respond slower than if you continued repeat the same task. The same is true for mistakes, more errors are committed when switching. However, switching cost can be reduced by using pre-stimulus cues and predictable sequences, such as those cues you see during quick-time events. On a sidenote, it would be interesting to see an experimenter-designed platform videogame using the switching and pre-stimulus cue paradigms. A good question from the authors is whether there any cost switching differences between non-gamers and gamers. Such differences could be due to videogame experience, processing speed or executive control processes. The authors noted some methodological observations about intertrial intervals where earlier studies with short ones (i.e. 150 milliseconds) found gamer experience differences. Thus, the authors would like to see if those gamer differences would show with longer interterial intervals (i.e. 1500milliseconds), no pre-stimulus cues and unpredictable sequences.
Distractor filtering was thought to be greater among gamers as they must filter considerable amount of irrelevant visual information in videogames. However, past studies have failed to find such differences in high or low perceptual load (i.e. the amount of irrelevant visual information presented). The authors examined another aspect of distractor filtering through trial history. This trial history is examined by the performance level of one trial in relation to its previous ones. For example, if the current trial is the same as the previous or if the current trial is very dissimilar from the previous ones, the latter would be harder than the former because it is a distractor.
Finally, logical memory was hypothesized not to be affected by videogame experience. This was meant to address concerns posed by Walter Boot and colleagues last year about global differences between gamers and non-gamers, such as intelligence and motivation. In particular, whether gamer participants would be suspicious if their gaming abilities was the focus of the experiments. The authors hypothesized that no differences in a logical memory task would suggest that differences in the task switching cognitive task would be due to specific cognitive processes.
Participants: 44 undergraduate students from the University of California, Berkeley participated. Participants were pre-screened for videogame experiences. They were carefully pre-screened so as not to arouse their suspicions about the study’s goals. If they were suspicious, according to Walter Boot, the possibility of increased motivation would become a possible third variable in the results, something that earlier studies had not taken into account. Participants were classified into the gamer group if they played videogames for 6 hours or more per week, primarily FPS or action and that they ranked their expertise on FPS or action on a level of 5 or higher. The number of gamer participants is 23, average age is 20.8 year and are all men. Non-gamers were those who played less than 2 hours per week on FPS and action and whose self-reported expertise level is lower than 2. The number of non-gamer participants is 21, average age is 22.5, and 13 are women and 8 are men.
The authors did note about gender difference in their sample and argued that prior studies with equal numbers of female and male participants reported finding no gender differences in their results. Therefore, the authors concluded that differences found in their results are due to videogame experience rather than gender differences.
There are a lot of details, but I’ll break out down to what a single participant experienced. First, they were shown a white circle in the middle of the screen between 1,800-2,200 milliseconds. This is the intertrial interval, the waiting period. Then they are shown three sets of arrows (see below for one of them). They were given prior instructions that if they see blue arrows, then they must indicate by key press the direction of the center arrow and ignore the other arrows (they’re distractors). So if a blue center arrow is pointing left, then press the ‘left’ key. If they see yellow arrows, then they must indicated by the opposite direction of the center arrow. So if a yellow center arrow is pointing right, then press the ‘left’ key. The other arrows could be presented congruently (those on the left side) or incongruently (those on the right side), so you’d imagined that it’s harder on response time for the right side. These arrows are shown for 1,300 milliseconds.
The presentation order was made that there would be equal numbers of pro-response repeated (two blue trials), pro-response switched (yellow trial and then blue trial), anti-response repeated (two yellow trials) and anti-response switched (blue trial and then yellow trial). The same is true for the distractor arrows.
The participant was given an eight-trial practice round. Following that, they have four blocks of 85 trials.
They conducted a mixed-model 2 (response type) X 2 (gamer group) X 2 (trial history) ANOVA, the former two as between-subjects and the last as the within-subjects factor. They removed outliers in response times and mistakes from the ANOVA. Their results yielded expected results, a main effect for response type where the blue arrows is faster than the yellow arrows, a main effect for trial history where repeating the blue arrow trials has the fastest response time. No main differences between the non-gamer vs. gamer group. An interaction effect was found for response type and trial history where switching from yellow arrows to blue arrows resulted in slower response time.
A three-way interaction was found in that non-gamers had a slower response time switching from yellow arrows to blue arrows, but a faster response time for switching from blue arrows to yellow arrows. The gamer group had similar results, but they were much faster (37 vs. 16 milliseconds) than the non-gamer group (63 vs. 9 milliseconds). The authors wrote that the gamer group had a more symmetric pattern, so I understand that the gamers’ 37 vs. 16 milliseconds are more symmetric than the non-gamers more asymmetric 63 vs. 9 milliseconds response.
The authors conducted the same mixed-model ANOVA with the errors and outliers included. They found similar results in the first ANOVA where there is more accuracy for the blue arrows than yellow arrows. More mistakes were made for switching from yellow arrows to blue arrows than switching blue arrows to yellow arrows. No gamer group differences were found or any three-way interactions.
A third ANOVA was conducted to examine the flanker effect, the distracter arrows that either followed or don’t follow the center arrow’s direction. They found what was expected in that participants had a faster response time for the distracter arrows that followed than those that don’t follow. An interaction effect was found in that a larger effect for trials following a congruent trial than for trials following incongruent trial. No effects were found for the gamer groups.
Before going to the discussion section, the authors conducted another experiment with the same participants testing for differences in logical memory. The participants listened to two stories that lasted for 30 seconds each. They were told to recall as many details as possible and they can take as long as they want. After they recalled details for the second story, the second story was repeated again and the participants then again recalled the details. The experimenter checks off what details the participants remembered.
What their results revealed is that there were no differences in recall between groups. In effect, the difference found in the task switching experiment is unlikely attributed to intelligence, motivation or suspiciousness, although the authors noted that the recall task says little about gamers’ performance.
The take home message is that without pre-stimulus cues, predictable sequences and longer intertrial intervals, gamers’ cognitive performance are weaker than previously reported. This means alternative explanations are afoot, said the authors, in regards to those previous studies. About the symmetric responses, the authors contend that gamers have a mechanistic shift in performance where they switch tasks more flexibly than non-gamers. The authors posed this question of what underlying cognitive mechanisms may differ in the way that gamers handle switch tasks. The authors offered several reasons: gamers may “activate each task representation less strongly”, resulting in symmetry response time, but overall slower response time. However, this is discounted by their current findings. Another reason is that gamers are not following a pattern for trial history, so their response times for two blue arrows or two yellow arrows are slower than the non-gamers which the authors pointed out in their results section.
In regards to the distractor arrows, action videogames does not improve visual information filtering. I might think that it could worsen if we look at others’ research, say Kira Bailey’s research on attention.
Addressing to Walter Boot et al.’s methodological concerns, their findings suggest that differences are not due to intelligence or motivation, but due to gamer-expertise and executive control system tested by the study’s task.